Accuracy in the outcomes and assessment information set (OASIS): Results of a video simulation

Abstract

There is little information regarding the accuracy of the Outcomes and Assessment Information Set (OASIS), the patient assessment tool mandated for use in Medicare-funded home health care. The purposes of this study were to evaluate the accuracy of OASIS completion by home health nurses and rehabilitation therapists, to compare responses of nurses and therapists, and to determine whether dispersion of answers would affect the home health resource group (HHRG) to which patients were assigned for Medicare home health care payments to agencies. Using a video simulation of admission and discharge visits, 436 clinicians from 29 Ohio home health care agencies scored selected OASIS items. Although the majority of the items were rated accurately, discrepancies were found between clinician responses and the "correct" answer on several items. Nurses and therapists provided similar ratings on most items studied, but for most cases in which discrepancies were found, nurses were more likely to agree with the "correct" answer. Discrepancies most often led to patients being assigned to lower-payment HHRGs. Continued monitoring of OASIS data collection accuracy is recommended to maximize the value of the OASIS instrument in home health care research, practice, and policy.

Racial/Ethnic Disparities in Home Health Care: Charting a Course for Future Research

“…Assessment systems that are also used for payment can generate incentives for assessors to overstate impairments at admission in order to increase the agency's payment while decreasing impairment at discharge to expedite treatment termination. However, research shows that the minimal errors in OASIS scoring result in an underestimate of patient impairment at admission and there is greater accuracy at discharge than admission (Madigan et al, 2003). Since patients can be continued on home health care indefinitely as long as the documentation supports the need, there is little incentive for agencies to understate impairment at discharge.…”

“…Davitt & Choi, 2008a;Madigan, Tullai-McGuinness, & Fortinsky, 2003 Control * This is not an exhaustive list of indicators or hypotheses, rather merely an exemplary list of some of the key indicators for each dimension. * * Direction is predicted where it is supported by the evidence.…”

“…Also, reliability studies and qualitative research on home care assessment processes indicate that research should control for the discipline of the staff assessing the patient (e.g., nurse or therapist). Different assessment approaches may impact the illness severity rating based on whether observations of actual performance or patient self-report of functional abilities are obtained (Davitt & Choi, 2008a;Madigan, Tullai-McGuinness, & Fortinsky, 2003). …”

The Outcome and Assessment Information Set (OASIS): A Review of Validity and Reliability

“…It is important to consider these purposes when critiquing the reliability and validity of OASIS. Home care clinicians can complete the OASIS to benefit the home care agency in reimbursements or outcome indicators, compromising the reliability and validity of the tool and its value in understanding quality and patient outcomes (Davitt, 2009; Davitt & Choi, 2008; Madigan, Tullai-McGuinness, Fortinsky, 2003). According to CMS, upcoding, or overstating the severity of a patient’s health status, accounted for 11.78% of the change in case-mix between 2000 and 2008 (Davitt & Kaye, 2010; U.S.…”

“…Both expert and clinician input and statistical procedures were employed to “measure and risk adjust patient outcomes in home health” (Shaughnessy, Crisler & Schlenker, 1998, p. 64). The research team that developed the OASIS reported interrater reliability kappas ranging from .50 to 1 on functional variables, .6 for dyspnea and pain, and .79 to 1 for behavioral items (Shaughnessy et al, 1994, as cited in Madigan et al, 2003). A second study by the developers of the OASIS found acceptable reliability levels with most items (71%) having weighted kappas of at least .60 (Schlenker, Powell, Goodrich, & Kaehny, 2000, as cited in Hittle et al, 2003).…”